HAL Id: hal-03281012https://hal.archives-ouvertes.fr/hal-03281012

Submitted on 7 Jul 2021

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Addressing Organic Electrochemical Transistors forNeurosensing and Neuromorphic Sensing

Mahdi Ghazal, Thomas Dargent, Sébastien Pecqueur, Fabien Alibart

To cite this version:Mahdi Ghazal, Thomas Dargent, Sébastien Pecqueur, Fabien Alibart. Addressing Organic Electro-chemical Transistors for Neurosensing and Neuromorphic Sensing. IEEE SENSORS 2019, Oct 2019,Montréal, Canada. 10.1109/SENSORS43011.2019.8956648. hal-03281012

Addressing Organic Electrochemical Transistors for

Neurosensing and Neuromorphic Sensing

Mahdi Ghazal

École Supérieure

d’Ingénieurs en

Électrotechnique et

Électronique, Paris, France

Institut d’Électronique,

Microélectronique et

Nanotechnologie,

Villeneuve d’Ascq, France

Thomas Dargent

Institut d’Électronique,

Microélectronique et

Nanotechnologie,

Villeneuve d’Ascq, France

Sebastien Pecqueur*

Institut d’Électronique,

Microélectronique et

Nanotechnologie,

Villeneuve d’Ascq, France

sebastien.pecqueur@iemn.u

niv-lille1.fr

Fabien Alibart

Institut d’Électronique,

Microélectronique et

Nanotechnologie,

Villeneuve d’Ascq, France

University of Sherbrooke,

QC, Canada

Abstract—We report on the comparison of two different

driving circuits for addressing micro-fabricated organic

electrochemical transistors of different channel resistances and

transconductance, aiming for neuromorphic sensing. While the

Current-Voltage converter observed faster transients, the

Wheatstone bridge configuration offers more versatility towards

higher resistance materials. Both circuits shows different assets

very encouraging for further practical application.

Keywords—OECT; sensing circuit; signal read-out; current-

voltage converter; Wheatstone bridge

I. INTRODUCTION

In the quest of monitoring living matter activities,

recording of extracellular potentials in electroactive cells

represents a major challenge. For instance, this sensing

approach can open various application fields, from health

diagnosisfor (e.g. brain disease) to bioelectronics

communication (i.e. Brain Machine Interfaces). Lately, the

performance of organic bioelectronics, particularly organic

electrochemical transistors (OECTs), has received lots of

attention for extracellular recordings since they exhibit a better

coupling with the cells, improve biocompatibility and better

signal-to-noise ratio since they can be considered as active

elements.[1] In addition, OECTs have already been used in

various biomolecular sensors applications to measure

glucose,[2] dopamine,[3] lactate,[4] and bacteria.[5]

Nevertheless, OECT’s applications for monitoring bio-

electrochemical activity in vivo or in vitro [6] or to implement

ionic sensor arrays are still at an early stage and integrating

such devices into large systems required addtionnal research

efforts.

The operation mechanism of a PEDOT:PSS OECT is

primarily based on the electrochemical doping/de-doping

processes when ions are injected/removed from the active

material by applying an electric field on the gate terminal. Ions

modulate the electrical conductivity of the semiconductor

organic material by oxidation/reduction.[7] The resulting

doping level can then be monitored by applying a source-drain

voltage bias to promote a current drift across the channel.

For instance, when recording electrical activity from.in in-

vitro neural cells cultures,[8] electroactive cells can be

considered as a power supply in the ionic circuit that produce

transient voltages that will be translated into modulation in the

output drain current. Also, measuring the transient response of

OECTs to pulse of voltages have been demonstrated as a

promising route for neuromorphic sensing and neuromorphic

computing giving access to a larger set of information and

Hence, OECTs as a sensor produces a change in its electrical

conductivity to indicate a change in its chemical or biological

environment. To record this change in electrical property, the

current output signal needs to be conditioned by an analog

circuit. Such additional sensing circuitry required further

analysis since it will be an important element of the overall

measurement system in any applications. Notably, such

additional circuitry need to be adapted to OECTs specificity

such as dynamic range of operation and variability among

devices. In this paper, we explore different analog circuits for

sensing electrical signals from OECT devices.

II. OECT ELECTRICAL CHARACTERISTICS

In the output curves of the OECTs (displayed in Fig. 1(a)),

the drain current increases linearly with the drain voltage,

maintaining an Ohmic behavior from 0 to 1 V. The

characterized devices have shown different channel resistance

values from 100 Ω to 10 GΩ as shown in the histogram in

Fig. 1(a). In the transfer characteristics of the OECTs

(displayed in Fig. 1(b)), we observed an averaged

transconductance of 1 µS for the 40 measured OECTs: these

wide spreads in device performances gives us the opportunity

to probe the capability for our addressing circuits for various

electrical properties of the OECT.

III. OECT SENSOR CIRCUITS

Here, the electronic setups are implemented to measure signals

from a single OECT by using an analog measurement device

(e.g Oscilloscope). The working mechanism of these electronic

circuits relies on converting the drain current or the resistance

of the OECT into voltage, signal amplification, and filtering.

One setup considers OECTs as a current source sensors,

based on the transimpedance circuit configuration.

A second setup considers OECT as a resistive sensor due

to its Ohmic behavior (Fig. 1(a)) based on a Wheatstone

bridge circuit.

A. OECT with a Transimpedance Configuration

The working mechanism of this recording circuitry is shown in Fig. 2. Any electrical activity, coming from the electrogenic cells sensed by OECT as current modulations, is converted to voltage changes through the I/V convertor with a feedback resistor of 10 kΩ at the first stage. To remove the offset signal (at constant gate voltage) and the noise of the low frequency bio-chemical signals, the resulted voltage is filtered by a 3rd-order Butterworth high-pass filter with cutoff frequency Fl = 1 Hz (with -60dB/decade of attenuation below Fl). The signal is then amplified by 50X at the third stage and then it is filtered by 3rd-order Butterworth low-pass filter with cutoff frequency Fh = 10 kHz (with -60dB/decade of attenuation beyond Fh) to limit the noise bandwidth (Fig. 2).

B. OECT with a Wheatstone Bridge Configuration

Wheatstone bridge circuits are commonly used for differential

resistance measurements. Due to its observed ohmic behavior,

OECTs can be considered as resistive sensors implementable in

a Wheatstone bridge with other load resistors (R) as shown in

Fig. 3. Any electrical activity in the electrolyte coming from the

electrogenic cells sensed by the OECT will cause small

variation (R in the resistance of the OECT (𝑅𝑂𝐸𝐶𝑇) that will

induce a change in the output signal of the Wheatstone bridge

described by (1).

If the load resistances are calibrated manually via

potentiometers such that they equal the one of the OECT

(𝑅𝑂𝐸𝐶𝑇) at rest, the output signal can be described by (2).

Where VCC is the low voltage supply (100 mV). Therefore, the

electrical activities of the cells will modulate the output voltage

of the Wheatstone bridge that can be monitored later.

The resulted output voltage of the Wheatstone bridge is

then amplified by a 100X differential amplifier. The amplified

signal is then passed through high- and low-pass filters

separated by a buffer. The filters configurations are

implemented with the same type and design than the ones

previously implemented in the Transimpedance configuration.

Fig. 3.Schematic of the OECT with the Wheatstone bridge configuration

IV. EXPERIMENTS AND RESULTS

For each circuit configuration, a PCB circuit has been

fabricated. The 42-OECT chip was interfaced with the PCB by

wire bonding on a CLCC (Ceramic Leadless Chip Carrier)

inserted in a PLCC (Plastic Leaded Chip Carrier) socket on the

board (Fig. 4). In the Transimpedance configuration PCB, the

This work was funded by the European Commission under the

ERC-2017-COG program, IONOS project GA773228.

Fig. 1. Electrical characteristics of OECTs: (a) current-voltage output curve; (b) transfer characteristics curve

Fig. 2 Schematic of the OECT with the I-V converter circuit

𝑉𝑜𝑢𝑡 = (1

2−

𝑅𝑂𝐸𝐶𝑇 + ∆𝑅

𝑅 + (𝑅𝑂𝐸𝐶𝑇 + ∆𝑅))𝑉𝑠

(1)

Fig. 1 Characteristics of the OECTs used to assess the circuits: (a) Output characteristics of a population displaying an Ohmic behavior in the applied voltage

range (inset: histogram of their channel resistances). (b) Transfer characteristics of the same population (inset: histogram of their transconductance)

𝑉𝑜𝑢𝑡 = (∆𝑅

𝑅 +∆𝑅2

)𝑉𝐶𝐶4

(2)

OECTs drain-source channels are biased at the same time by a

100 mV biasing voltage and each OECT output can be selected

to connect the input of the circuit by a jumping wire. In the

Wheatstone bridge configuration PCB, the source and the drain

of each OECT can be selected and connected to the Wheatstone

bridge circuit by jumping wires where each OECT will be

biased from the alimentation input of the Wheatstone bridge

(Fig. 4).

Despite no particular shielding, both electronic circuits

achieved a noise level (root mean square) in the range of 10-

20 mV which was low enough to monitor clearly the electrical

spikes at the output of these circuits as shown in Fig. 5.

Fig. 4. Picture of the Wheatstone bridge configuration circuit with the

Ag/AgCl gate electrode dipped in the PBS electrolyte.

The experiments were done for investigating the response of the

OECT output to transient signals (e.g. voltage pulses applied to

the gate) in Phosphate-Buffered Saline (PBS). The fix drain

potential (and the Wheatstone bridge circuit input bias for the

Wheatstone bridge configuration) was set to 100 mV and the

output of each circuit was measured continuously at a pulsed

gate voltage Vgate = 100 mV over a 0.5 s (2 Hz frequency)

period with 1 ms pulse width. These experiments were tested

on four different OECTs, from four different channel resistivity

values as they were categorized in the histogram in Fig. 1a. As

shown in the graphs in Fig. 5. Both circuits successfully

recorded the simulated gate input signal, with the offset DC

output signal being removed in both cases. Comparing both

circuits by referring to Fig. 5b and Fig. 5c, the value of the

output of the I-V converter circuit is directly related to the

resistivity of the OECT channel. For lower resistivity values,

we got higher drain currents and thus higher output voltage.

However, for the Wheatstone bridge circuit, in addition to the

resistivity of the OECT, the output voltage depends also on the

values of the fixed load resistances (as shown in (2)). Hence,

Wheatstone bridge circuit is still sensitive for OECTs with high

resistance values while on the other hand, the time of the charge

and discharge is noticeably slower than the I-V converter output

response.

V. CONCLUSION

This paper reports the implementation of two electronic circuits

for addressing OECT for neuromorphic sensing applications.

Electrical characterizations of the OECTs also been carried out

to show the several electrical properties of the devices before

testing them with the circuits. These circuits showed successful

response after their tested on a simulated signal of the cells

promising that it will provide effective results for real

biological measurements for electrogenic cells.

Acknowledgements

We acknowledge financial supports from the EU: ERC-2017-

COG project IONOS (# GA 773228). We thank the French

National Nanofabrication Network RENATECH for financial

support of the IEMN clean-room. We thank also the IEMN

cleanroom staff of their advices and support

Fig. 5. Response of the output voltage on the simulated signals of electrogenic cells: (a) of the I-V converter circuit, (b) zoomed shot of one spike at the

output of the I-V converter circuit, (c) zoomed shot of one spike at the output

Wheatstone bridge circuits.

REFERENCES

[1] D. Khodagholy et al., « High transconductance organic

electrochemical transistors », Nat. Commun., vol. 4, p.

2133, july 2013.

[2] S. V. Rajaraman, S. Choi, M. A. McClain, J. de R. Ross,

M. C. LaPlaca, et M. G. Allen, « Metal-Transfer-

Micromolded Three-Dimensional Microelectrode Arrays

for in-vitro Brain-Slice Recordings », J.

Microelectromechanical Syst., vol. 20, p. 396‑409, 2011.

[3] J. P. Spatz et B. Geiger, « Molecular engineering of

cellular environments: cell adhesion to nano-digital

surfaces », Methods Cell Biol., vol. 83, p. 89‑111, 2007.

[4] D. Khodagholy et al., « Organic electrochemical

transistor incorporating an ionogel as a solid state

electrolyte for lactate sensing », J. Mater. Chem., vol. 22,

no 10, p. 4440‑4443, feb. 2012.

[5] C. Liao, M. Zhang, L. Niu, Z. Zheng, et F. Yan,

« Organic electrochemical transistors with graphene-

modified gate electrodes for highly sensitive and selective

dopamine sensors », J. Mater. Chem. B, vol. 2, no 2, p.

191‑200, dec. 2013.

[6] « Organic Electrochemical Transistor Arrays for In Vitro

Electrophysiology Monitoring of 2D and 3D Cardiac

Tissues ».

[7] C. Liao, M. Zhang, M. Y. Yao, T. Hua, L. Li, et F. Yan,

« Flexible Organic Electronics in Biology: Materials and

Devices », Adv. Mater. Deerfield Beach Fla, vol. 27, no

46, p. 7493‑7527, dec. 2015.

[8] M. E. Spira et A. Hai, « Multi-electrode array

technologies for neuroscience and cardiology », Nat.

Nanotechnol., vol. 8, no 2, p. 83‑94, feb. 2013.

[9] S. Pecqueur, D. Guérin, D. Vuillaume, F. Alibart, Cation

Discrimination in Organic Electrochemical Transistors by

Dual Frequency Sensing, Org. Electron. 57 (2018), 232-238.